of prairie voles in outdoor enclosures, Mahady and Wolff
(2002) replaced resident males with unfamiliar males every
10 days. They reported that 7 of 33 (21%) nulliparous fe-
males did not conceive during the study, but whether this
was due to pregnancy failure or disruption of pairbonding
in this monogamous species is not known. Production of
second litters and breeding by parous females were not af-
fected by exposure to strange males. Certainly, more field
studies are needed to test the validity of the Bruce effect or
pregnancy disruption hypothesis as a counterstrategy to in-
fanticide, but at least these two field studies with two Mi-
crotusspecies do not provide strong support that pregnancy
disruption occurs regularly or is an adaptive response to
exposure to strange males, at least in this taxon.
Among sciurids, most takeovers by male Alpine marmots
(62%, n 21) occur after the mating period or before the
end of lactation (Hackländer and Arnold 1999). Interest-
ingly, female breeding is reduced after these male takeovers
despite clear signs of pregnancy early in the season, and fe-
males failing to reproduce right after these takeovers in-
crease their chance of breeding in the following year (Hack-
länder and Arnold 1999). Nonetheless, male takeovers in
other populations of Alpine marmots seem to occur mostly
(75%, n 20) when juveniles are already born (King and
Allainé 2002). Taken together, these field studies provide
only moderate support for the idea that pregnancy disrup-
tion is a strategy to prevent losses to infanticide. Moreover,
predators and other potentially stressful factors also may
cause pregnancy disruptions in female rodents (de Catan-
zaro and MacNiven 1992), which suggests that pregnancy
disruption may indeed be part of a more general strategy
to prevent the waste of energy in producing offspring likely
to be lost.
Finally, socially subordinate individuals may suppress
breeding as a strategy to avoid wasting energy and resources
on litters that are likely to be eliminated by more dominant
females within the group (Agrell et al. 1998). Such may be
partially the case of subordinate females of Alpine marmots
that achieve copulations and become pregnant within their
social units, but only the dominant females give birth (Hack-
länder and Arnold 1999; King and Allainé 2002). Formal
phylogenetic analyses may shed light onto the evolutionary
relationships between the occurrence of within-group in-
fanticide, social living, and breeding suppression.Concluding RemarksFunctions of infanticide
The functional significance of infanticide in rodents is com-
plex and cannot be explained by any one single hypothesis.
Each hypothesis has its own assumptions, predictions, and
tests (fig. 23.1; table 23.2). In some species individuals ob-
tain nutritional benefits from infanticide (table 23.2). In
some cases, nutritional benefits are gained by females (e.g.,Nonparental Infanticide 277Figure 23.1 Key conditions and the subsequent benefits from infanticidal behavior reported to occur in male and female rodents. Not all benefits are equally well
supported (see text).